[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"extension-skill-huggingface-hf-mcp-de":3,"guides-for-huggingface-hf-mcp":710,"similar-k17c8k3pq66ym9rfgr6vbnkk3986mdsx-de":711},{"_creationTime":4,"_id":5,"children":6,"community":7,"display":9,"evaluation":15,"identity":245,"isFallback":242,"parentExtension":249,"providers":250,"relations":255,"repo":257,"tags":707,"workflow":708},1778690773482.4863,"k17c8k3pq66ym9rfgr6vbnkk3986mdsx",[],{"reviewCount":8},0,{"description":10,"installMethods":11,"name":13,"sourceUrl":14},"Use Hugging Face Hub via MCP server tools. Search models, datasets, Spaces, papers. Get repo details, fetch documentation, run compute jobs, and use Gradio Spaces as AI tools. Available when connected to the HF MCP server.",{"claudeCode":12},"huggingface/skills","hf-mcp","https://github.com/huggingface/skills",{"_creationTime":16,"_id":17,"extensionId":5,"locale":18,"result":19,"trustSignals":226,"workflow":243},1778691205845.1924,"kn7b1er5w7wgqpg8x5b35dr2cd86mtbd","en",{"checks":20,"evaluatedAt":193,"extensionSummary":194,"features":195,"nonGoals":201,"promptVersionExtension":205,"promptVersionScoring":206,"purpose":207,"rationale":208,"score":209,"summary":210,"tags":211,"targetMarket":218,"tier":219,"useCases":220},[21,26,29,32,36,39,43,47,50,53,57,61,65,69,72,75,78,81,84,87,91,95,99,103,107,110,114,117,121,124,127,130,133,136,139,143,147,150,154,158,161,164,167,170,174,177,180,183,186,190],{"category":22,"check":23,"severity":24,"summary":25},"Practical Utility","Problem relevance","pass","The description clearly identifies the problem of connecting AI assistants to the Hugging Face Hub and lists specific tasks like searching models, datasets, and running compute jobs.",{"category":22,"check":27,"severity":24,"summary":28},"Unique selling proposition","The skill provides a set of specialized tools for interacting with the Hugging Face Hub, going beyond basic API calls by offering structured search, job management, and Gradio space integration.",{"category":22,"check":30,"severity":24,"summary":31},"Production readiness","The skill provides a comprehensive set of tools for interacting with Hugging Face Hub, covering search, retrieval, job execution, and dynamic space interaction, enabling a full lifecycle for many ML workflows.",{"category":33,"check":34,"severity":24,"summary":35},"Scope","Single responsibility principle","The skill is focused on interacting with the Hugging Face Hub and its associated services, with a coherent set of tools for managing models, datasets, jobs, and spaces.",{"category":33,"check":37,"severity":24,"summary":38},"Description quality","The displayed description accurately reflects the capabilities outlined in the SKILL.md file.",{"category":40,"check":41,"severity":24,"summary":42},"Invocation","Scoped tools","The skill exposes narrowly scoped tools like `model_search`, `dataset_search`, and `hf_jobs`, rather than a single generalist execution tool.",{"category":44,"check":45,"severity":24,"summary":46},"Documentation","Configuration & parameter reference","All parameters for the exposed tools are documented within the SKILL.md file, including descriptions and examples.",{"category":33,"check":48,"severity":24,"summary":49},"Tool naming","Tool names like `model_search`, `dataset_search`, and `hub_repo_details` are descriptive and adhere to the verb-noun convention.",{"category":33,"check":51,"severity":24,"summary":52},"Minimal I/O surface","Tool inputs are well-defined structured arguments, and outputs seem focused on the requested information without excessive extraneous data.",{"category":54,"check":55,"severity":24,"summary":56},"License","License usability","The bundled LICENSE file is the Apache License 2.0, which is a permissive open-source license.",{"category":58,"check":59,"severity":24,"summary":60},"Maintenance","Commit recency","The last commit was on 2026-05-12, which is within the last 3 months.",{"category":58,"check":62,"severity":63,"summary":64},"Dependency Management","not_applicable","The skill does not appear to use third-party dependencies directly within its SKILL.md; operations rely on the underlying MCP server or Hugging Face CLI.",{"category":66,"check":67,"severity":24,"summary":68},"Security","Secret Management","Secrets like HF_TOKEN are handled via environment variables and passed securely to `hf_jobs`, with no evidence of secrets being logged or exposed.",{"category":66,"check":70,"severity":24,"summary":71},"Injection","The skill relies on well-defined tool interfaces and does not appear to execute arbitrary code or load untrusted external content as instructions.",{"category":66,"check":73,"severity":24,"summary":74},"Transitive Supply-Chain Grenades","The skill operates within the defined bundle and relies on the MCP server; there's no evidence of runtime downloads or execution of external code.",{"category":66,"check":76,"severity":24,"summary":77},"Sandbox Isolation","The skill interacts with the Hugging Face Hub via defined tools, and there's no indication of it modifying files outside its intended scope or project folder.",{"category":66,"check":79,"severity":24,"summary":80},"Sandbox escape primitives","No detached processes or retry loops around denied tool calls were observed in the skill's definition.",{"category":66,"check":82,"severity":24,"summary":83},"Data Exfiltration","The skill's outbound calls are to documented Hugging Face Hub endpoints, and sensitive data like tokens are handled securely.",{"category":66,"check":85,"severity":24,"summary":86},"Hidden Text Tricks","The bundled content appears to be free of hidden steering tricks or obfuscated instructions.",{"category":88,"check":89,"severity":24,"summary":90},"Hooks","Opaque code execution","The skill definition uses standard tool definitions and does not employ obfuscated code execution techniques like base64 payloads or eval.",{"category":92,"check":93,"severity":24,"summary":94},"Portability","Structural Assumption","The skill makes no assumptions about the user's project file structure, interacting with the Hugging Face Hub via its API.",{"category":96,"check":97,"severity":24,"summary":98},"Trust","Issues Attention","In the last 90 days, 22 issues were opened and 6 were closed, indicating active maintenance and response.",{"category":100,"check":101,"severity":24,"summary":102},"Versioning","Release Management","The `pushedAt` date indicates recent activity, and the `hf-cli` skill is the primary interface, which is likely versioned separately by its own mechanisms.",{"category":104,"check":105,"severity":24,"summary":106},"Execution","Validation","Tool arguments are well-defined in the schema, implying validation and sanitization by the underlying MCP server or tool definitions.",{"category":66,"check":108,"severity":24,"summary":109},"Unguarded Destructive Operations","Destructive operations like job creation or model uploads are guarded by explicit tool calls and potentially require user confirmation through the agent.",{"category":111,"check":112,"severity":24,"summary":113},"Code Execution","Error Handling","The tool definitions imply structured error handling by the MCP server, and explicit examples in SKILL.md guide the user on recovery steps.",{"category":111,"check":115,"severity":63,"summary":116},"Logging","This skill primarily interacts with external services via defined tools, and detailed logging is expected to be handled by the MCP server or the Hugging Face CLI itself.",{"category":118,"check":119,"severity":63,"summary":120},"Compliance","GDPR","The skill operates on Hugging Face Hub resources and does not appear to handle personal data directly.",{"category":118,"check":122,"severity":24,"summary":123},"Target market","The skill interacts with the Hugging Face Hub, which is a global platform, and there are no regional restrictions indicated.",{"category":92,"check":125,"severity":24,"summary":126},"Runtime stability","The skill relies on standard tools and the Hugging Face Hub API, making it portable across different environments that can access these services.",{"category":44,"check":128,"severity":24,"summary":129},"README","The README provides a good overview of Hugging Face skills, installation instructions, and details on available skills.",{"category":33,"check":131,"severity":24,"summary":132},"Tool surface size","The skill exposes a reasonable number of tools (around 12 distinct operations) for its domain.",{"category":40,"check":134,"severity":24,"summary":135},"Overlapping near-synonym tools","The tool names are distinct and cover specific functionalities, avoiding near-synonym overlap.",{"category":44,"check":137,"severity":24,"summary":138},"Phantom features","All advertised capabilities in the SKILL.md and README are supported by concrete tools and examples.",{"category":140,"check":141,"severity":24,"summary":142},"Install","Installation instruction","The README provides clear installation instructions for multiple agents (Claude Code, Codex, Gemini CLI, Cursor) and includes copy-pasteable examples.",{"category":144,"check":145,"severity":24,"summary":146},"Errors","Actionable error messages","Examples and tool descriptions imply actionable error reporting, guiding users on remediation steps for common issues like missing credentials or rate limits.",{"category":104,"check":148,"severity":63,"summary":149},"Pinned dependencies","The skill itself does not bundle scripts with dependencies; it relies on the MCP server and Hugging Face CLI which are managed externally.",{"category":33,"check":151,"severity":152,"summary":153},"Dry-run preview","info","While state-changing operations exist, a specific `--dry-run` flag is not explicitly documented for all tools; however, the agent can review intended actions before execution.",{"category":155,"check":156,"severity":152,"summary":157},"Protocol","Idempotent retry & timeouts","The underlying Hugging Face Hub API and MCP server are expected to handle retries and timeouts. The skill definition does not explicitly detail idempotency for mutations.",{"category":118,"check":159,"severity":24,"summary":160},"Telemetry opt-in","No explicit telemetry is mentioned as being collected by this skill; any such collection would likely be handled by the underlying Hugging Face CLI or MCP server, which typically requires opt-in.",{"category":40,"check":162,"severity":24,"summary":163},"Precise Purpose","The SKILL.md frontmatter and description clearly state the purpose (interacting with Hugging Face Hub) and provide specific use cases and boundaries.",{"category":40,"check":165,"severity":24,"summary":166},"Concise Frontmatter","The frontmatter is concise, clearly stating the core capability and providing a list of relevant examples and triggers without excessive length.",{"category":44,"check":168,"severity":24,"summary":169},"Concise Body","The SKILL.md is well-structured, using progressive disclosure for examples and keeping the core instructions concise.",{"category":171,"check":172,"severity":24,"summary":173},"Context","Progressive Disclosure","Examples are well-organized and illustrative, with longer procedures or complex details likely managed by the underlying tools or referenced documentation.",{"category":171,"check":175,"severity":63,"summary":176},"Forked exploration","The skill does not involve deep exploration or code review requiring a forked context; it directly calls defined tools.",{"category":22,"check":178,"severity":24,"summary":179},"Usage examples","The SKILL.md provides a comprehensive set of end-to-end examples for all major functionalities, demonstrating inputs, invocations, and expected outcomes.",{"category":22,"check":181,"severity":24,"summary":182},"Edge cases","The SKILL.md covers various failure modes and limitations, such as authentication issues, rate limits, and missing credentials, with suggested recovery steps.",{"category":92,"check":184,"severity":63,"summary":185},"Tool Fallback","The skill is designed to work with the Hugging Face MCP server and CLI, which are the primary intended targets; no fallback mechanism is necessary.",{"category":187,"check":188,"severity":24,"summary":189},"Safety","Halt on unexpected state","The skill guides the agent to halt and report on unexpected states, such as authentication failures or invalid inputs, ensuring a controlled workflow.",{"category":92,"check":191,"severity":24,"summary":192},"Cross-skill coupling","The skill is self-contained and focuses on Hugging Face Hub interactions; it does not implicitly rely on other skills and cross-links are not needed.",1778691205463,"This skill provides a set of tools to interact with the Hugging Face Hub, enabling users to search for models, datasets, and papers, get repository details, run compute jobs on GPUs, and utilize Gradio Spaces as AI tools. It relies on the Hugging Face MCP server for its operations.",[196,197,198,199,200],"Search models, datasets, Spaces, and papers on Hugging Face Hub","Fetch repository details and documentation","Run Python scripts and training jobs on cloud GPUs","Use Gradio Spaces as dynamic AI tools","Generate images and research topics",[202,203,204],"Managing local Python environments or dependencies directly","Providing a full IDE or code editor experience","Acting as a general-purpose shell or system administration tool","3.0.0","4.4.0","Connect AI assistants to the Hugging Face Hub, enabling tasks like finding models, managing datasets, and running cloud-based ML jobs.","The skill is well-documented, robustly implemented, and covers a broad range of Hugging Face Hub functionalities with clear examples and error handling. There are minor areas where explicit dry-run capabilities for all state-changing operations could be enhanced, but this does not detract from overall high quality.",98,"Excellent skill for interacting with Hugging Face Hub, providing comprehensive tools for models, datasets, and compute jobs.",[212,213,214,215,216,217],"huggingface","mcp","mlops","models","datasets","jobs","global","verified",[221,222,223,224,225],"Finding the best model for a specific task","Comparing models from different providers","Discovering datasets for machine learning projects","Using AI-powered tools hosted on Hugging Face Spaces","Running quick GPU jobs or training custom models",{"codeQuality":227,"collectedAt":229,"documentation":230,"maintenance":233,"security":239,"testCoverage":241},{"hasLockfile":228},false,1778691186203,{"descriptionLength":231,"readmeSize":232},222,9821,{"closedIssues90d":234,"forks":235,"hasChangelog":228,"openIssues90d":236,"pushedAt":237,"stars":238},6,663,4,1778593131000,10482,{"hasNpmPackage":228,"license":240,"smitheryVerified":228},"Apache-2.0",{"hasCi":242,"hasTests":228},true,{"updatedAt":244},1778691205845,{"basePath":246,"githubOwner":212,"githubRepo":247,"locale":18,"slug":13,"type":248},"hf-mcp/skills/hf-mcp","skills","skill",null,{"evaluate":251,"extract":253},{"promptVersionExtension":205,"promptVersionScoring":206,"score":209,"tags":252,"targetMarket":218,"tier":219},[212,213,214,215,216,217],{"commitSha":254},"HEAD",{"repoId":256},"kd72xwt5xnc0ktc4p7smzfcp3986m959",{"_creationTime":258,"_id":256,"identity":259,"providers":260,"workflow":703},1778689536128.5474,{"githubOwner":212,"githubRepo":247,"sourceUrl":14},{"classify":261,"discover":696,"github":699},{"commitSha":254,"extensions":262},[263,280,290,298,306,314,322,330,338,346,354,362,370,378,386,394,437,442,448,454,471,477,484,526,537,556,562,582,594,618,676],{"basePath":264,"description":265,"displayName":266,"installMethods":267,"rationale":268,"selectedPaths":269,"source":278,"sourceLanguage":18,"type":279},"","Agent Skills for AI/ML tasks including dataset creation, model training, evaluation, and research paper publishing on Hugging Face Hub","huggingface-skills",{"claudeCode":12},"marketplace.json at .claude-plugin/marketplace.json",[270,273,275],{"path":271,"priority":272},".claude-plugin/marketplace.json","mandatory",{"path":274,"priority":272},"README.md",{"path":276,"priority":277},"LICENSE","high","rule","marketplace",{"basePath":281,"description":282,"displayName":283,"installMethods":284,"rationale":285,"selectedPaths":286,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-llm-trainer","Train or fine-tune language models using TRL on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes hardware selection, cost estimation, Trackio monitoring, and Hub persistence.","huggingface-llm-trainer",{"claudeCode":283},"inline plugin source from marketplace.json at skills/huggingface-llm-trainer",[287],{"path":288,"priority":277},"SKILL.md","plugin",{"basePath":291,"description":292,"displayName":293,"installMethods":294,"rationale":295,"selectedPaths":296,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-local-models","Use to select models to run locally with llama.cpp and GGUF on CPU, Mac Metal, CUDA, or ROCm. Covers finding GGUFs, quant selection, running servers, exact GGUF file lookup, conversion, and OpenAI-compatible local serving.","huggingface-local-models",{"claudeCode":293},"inline plugin source from marketplace.json at skills/huggingface-local-models",[297],{"path":288,"priority":277},{"basePath":299,"description":300,"displayName":301,"installMethods":302,"rationale":303,"selectedPaths":304,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-paper-publisher","Publish and manage research papers on Hugging Face Hub. Supports creating paper pages, linking papers to models/datasets, claiming authorship, and generating professional markdown-based research articles.","huggingface-paper-publisher",{"claudeCode":301},"inline plugin source from marketplace.json at skills/huggingface-paper-publisher",[305],{"path":288,"priority":277},{"basePath":307,"description":308,"displayName":309,"installMethods":310,"rationale":311,"selectedPaths":312,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-papers","Look up and read Hugging Face paper pages in markdown, and use the papers API for structured metadata like authors, linked models, datasets, Spaces, and media URLs when needed.","huggingface-papers",{"claudeCode":309},"inline plugin source from marketplace.json at skills/huggingface-papers",[313],{"path":288,"priority":277},{"basePath":315,"description":316,"displayName":317,"installMethods":318,"rationale":319,"selectedPaths":320,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-community-evals","Add and manage evaluation results in Hugging Face model cards. Supports extracting eval tables from README content, importing scores from Artificial Analysis API, and running custom evaluations with vLLM/lighteval.","huggingface-community-evals",{"claudeCode":317},"inline plugin source from marketplace.json at skills/huggingface-community-evals",[321],{"path":288,"priority":277},{"basePath":323,"description":324,"displayName":325,"installMethods":326,"rationale":327,"selectedPaths":328,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-best","Find the best AI model for any task by querying Hugging Face leaderboards and benchmarks. Recommends top models based on task type, hardware constraints, and benchmark scores.","huggingface-best",{"claudeCode":325},"inline plugin source from marketplace.json at skills/huggingface-best",[329],{"path":288,"priority":277},{"basePath":331,"description":332,"displayName":333,"installMethods":334,"rationale":335,"selectedPaths":336,"source":278,"sourceLanguage":18,"type":289},"skills/hf-cli","Execute Hugging Face Hub operations using the hf CLI. Download models/datasets, upload files, manage repos, and run cloud compute jobs.","hf-cli",{"claudeCode":333},"inline plugin source from marketplace.json at skills/hf-cli",[337],{"path":288,"priority":277},{"basePath":339,"description":340,"displayName":341,"installMethods":342,"rationale":343,"selectedPaths":344,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-trackio","Track and visualize ML training experiments with Trackio. Log metrics via Python API and retrieve them via CLI. Supports real-time dashboards synced to HF Spaces.","huggingface-trackio",{"claudeCode":341},"inline plugin source from marketplace.json at skills/huggingface-trackio",[345],{"path":288,"priority":277},{"basePath":347,"description":348,"displayName":349,"installMethods":350,"rationale":351,"selectedPaths":352,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-datasets","Explore, query, and extract data from any Hugging Face dataset using the Dataset Viewer REST API and npx tooling. Zero Python dependencies — covers split/config discovery, row pagination, text search, filtering, SQL via parquetlens, and dataset upload via CLI.","huggingface-datasets",{"claudeCode":349},"inline plugin source from marketplace.json at skills/huggingface-datasets",[353],{"path":288,"priority":277},{"basePath":355,"description":356,"displayName":357,"installMethods":358,"rationale":359,"selectedPaths":360,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-tool-builder","Build reusable scripts for Hugging Face Hub and API workflows. Useful for chaining API calls, enriching Hub metadata, or automating repeated tasks.","huggingface-tool-builder",{"claudeCode":357},"inline plugin source from marketplace.json at skills/huggingface-tool-builder",[361],{"path":288,"priority":277},{"basePath":363,"description":364,"displayName":365,"installMethods":366,"rationale":367,"selectedPaths":368,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-gradio","Build Gradio web UIs and demos in Python. Use when creating or editing Gradio apps, components, event listeners, layouts, or chatbots.","huggingface-gradio",{"claudeCode":365},"inline plugin source from marketplace.json at skills/huggingface-gradio",[369],{"path":288,"priority":277},{"basePath":371,"description":372,"displayName":373,"installMethods":374,"rationale":375,"selectedPaths":376,"source":278,"sourceLanguage":18,"type":289},"skills/transformers-js","Run state-of-the-art machine learning models directly in JavaScript/TypeScript for NLP, computer vision, audio processing, and multimodal tasks. Works in Node.js and browsers with WebGPU/WASM using Hugging Face models.","transformers-js",{"claudeCode":373},"inline plugin source from marketplace.json at skills/transformers-js",[377],{"path":288,"priority":277},{"basePath":379,"description":380,"displayName":381,"installMethods":382,"rationale":383,"selectedPaths":384,"source":278,"sourceLanguage":18,"type":289},"skills/huggingface-vision-trainer","Train and fine-tune object detection models (RTDETRv2, YOLOS, DETR and others) and image classification models (timm and transformers models — MobileNetV3, MobileViT, ResNet, ViT/DINOv3) using Transformers Trainer API on Hugging Face Jobs infrastructure or locally. Includes COCO dataset format support, Albumentations augmentation, mAP/mAR metrics, trackio tracking, hardware selection, and Hub persistence.","huggingface-vision-trainer",{"claudeCode":381},"inline plugin source from marketplace.json at skills/huggingface-vision-trainer",[385],{"path":288,"priority":277},{"basePath":387,"description":388,"displayName":389,"installMethods":390,"rationale":391,"selectedPaths":392,"source":278,"sourceLanguage":18,"type":289},"skills/train-sentence-transformers","Train or fine-tune sentence-transformers models across all three architectures: SentenceTransformer (bi-encoder embeddings), CrossEncoder (rerankers), and SparseEncoder (SPLADE). Covers loss selection, hard-negative mining, evaluators, distillation, LoRA, Matryoshka, and Hugging Face Hub publishing.","train-sentence-transformers",{"claudeCode":389},"inline plugin source from marketplace.json at skills/train-sentence-transformers",[393],{"path":288,"priority":277},{"basePath":264,"description":265,"displayName":266,"installMethods":395,"license":240,"rationale":396,"selectedPaths":397,"source":278,"sourceLanguage":18,"type":289},{"claudeCode":266},"plugin manifest at .claude-plugin/plugin.json",[398,400,401,402,405,407,409,411,413,415,417,419,421,423,425,427,429,431,433,435],{"path":399,"priority":272},".claude-plugin/plugin.json",{"path":274,"priority":272},{"path":276,"priority":277},{"path":403,"priority":404},"skills/hf-cli/SKILL.md","medium",{"path":406,"priority":404},"skills/huggingface-best/SKILL.md",{"path":408,"priority":404},"skills/huggingface-community-evals/SKILL.md",{"path":410,"priority":404},"skills/huggingface-datasets/SKILL.md",{"path":412,"priority":404},"skills/huggingface-gradio/SKILL.md",{"path":414,"priority":404},"skills/huggingface-llm-trainer/SKILL.md",{"path":416,"priority":404},"skills/huggingface-local-models/SKILL.md",{"path":418,"priority":404},"skills/huggingface-paper-publisher/SKILL.md",{"path":420,"priority":404},"skills/huggingface-papers/SKILL.md",{"path":422,"priority":404},"skills/huggingface-tool-builder/SKILL.md",{"path":424,"priority":404},"skills/huggingface-trackio/SKILL.md",{"path":426,"priority":404},"skills/huggingface-vision-trainer/SKILL.md",{"path":428,"priority":404},"skills/train-sentence-transformers/SKILL.md",{"path":430,"priority":404},"skills/transformers-js/SKILL.md",{"path":432,"priority":272},".mcp.json",{"path":434,"priority":277},"agents/AGENTS.md",{"path":436,"priority":277},".cursor-plugin/plugin.json",{"basePath":246,"description":10,"displayName":13,"installMethods":438,"rationale":439,"selectedPaths":440,"source":278,"sourceLanguage":18,"type":248},{"claudeCode":12},"SKILL.md frontmatter at hf-mcp/skills/hf-mcp/SKILL.md",[441],{"path":288,"priority":272},{"basePath":331,"description":443,"displayName":333,"installMethods":444,"rationale":445,"selectedPaths":446,"source":278,"sourceLanguage":18,"type":248},"Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing models, datasets, spaces, buckets, repos, papers, jobs, and more on the Hugging Face Hub. Use when: handling authentication; managing local cache; managing Hugging Face Buckets; running or scheduling jobs on Hugging Face infrastructure; managing Hugging Face repos; discussions and pull requests; browsing models, datasets and spaces; reading, searching, or browsing academic papers; managing collections; querying datasets; configuring spaces; setting up webhooks; or deploying and managing HF Inference Endpoints. Make sure to use this skill whenever the user mentions 'hf', 'huggingface', 'Hugging Face', 'huggingface-cli', or 'hugging face cli', or wants to do anything related to the Hugging Face ecosystem and to AI and ML in general. Also use for cloud storage needs like training checkpoints, data pipelines, or agent traces. Use even if the user doesn't explicitly ask for a CLI command. Replaces the deprecated `huggingface-cli`.",{"claudeCode":12},"SKILL.md frontmatter at skills/hf-cli/SKILL.md",[447],{"path":288,"priority":272},{"basePath":323,"description":449,"displayName":325,"installMethods":450,"rationale":451,"selectedPaths":452,"source":278,"sourceLanguage":18,"type":248},"Use when the user asks about finding the best, top, or recommended model for a task, wants to know what AI model to use, or wants to compare models by benchmark scores. Triggers on: \"best model for X\", \"what model should I use for\", \"top models for [task]\", \"which model runs on my laptop/machine/device\", \"recommend a model for\", \"what LLM should I use for\", \"compare models for\", \"what's state of the art for\", or any question about choosing an AI model for a specific use case. Always use this skill when the user wants model recommendations or comparisons, even if they don't explicitly mention HuggingFace or benchmarks.\n",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-best/SKILL.md",[453],{"path":288,"priority":272},{"basePath":315,"description":455,"displayName":317,"installMethods":456,"rationale":457,"selectedPaths":458,"source":278,"sourceLanguage":18,"type":248},"Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware. Use for backend selection, local GPU evals, and choosing between vLLM / Transformers / accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publication, or community-evals automation.",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-community-evals/SKILL.md",[459,460,463,465,467,469],{"path":288,"priority":272},{"path":461,"priority":462},"examples/.env.example","low",{"path":464,"priority":462},"examples/USAGE_EXAMPLES.md",{"path":466,"priority":462},"scripts/inspect_eval_uv.py",{"path":468,"priority":462},"scripts/inspect_vllm_uv.py",{"path":470,"priority":462},"scripts/lighteval_vllm_uv.py",{"basePath":347,"description":472,"displayName":349,"installMethods":473,"rationale":474,"selectedPaths":475,"source":278,"sourceLanguage":18,"type":248},"Use this skill for Hugging Face Dataset Viewer API workflows that fetch subset/split metadata, paginate rows, search text, apply filters, download parquet URLs, and read size or statistics.\r",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-datasets/SKILL.md",[476],{"path":288,"priority":272},{"basePath":363,"description":364,"displayName":365,"installMethods":478,"rationale":479,"selectedPaths":480,"source":278,"sourceLanguage":18,"type":248},{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-gradio/SKILL.md",[481,482],{"path":288,"priority":272},{"path":483,"priority":404},"examples.md",{"basePath":281,"description":485,"displayName":283,"installMethods":486,"rationale":487,"selectedPaths":488,"source":278,"sourceLanguage":18,"type":248},"Train or fine-tune language and vision models using TRL (Transformer Reinforcement Learning) or Unsloth with Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, model selection/leaderboards and model persistence. Use for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-llm-trainer/SKILL.md",[489,490,492,494,496,498,500,502,504,506,508,510,512,514,516,518,520,522,524],{"path":288,"priority":272},{"path":491,"priority":404},"references/gguf_conversion.md",{"path":493,"priority":404},"references/hardware_guide.md",{"path":495,"priority":404},"references/hub_saving.md",{"path":497,"priority":404},"references/local_training_macos.md",{"path":499,"priority":404},"references/reliability_principles.md",{"path":501,"priority":404},"references/trackio_guide.md",{"path":503,"priority":404},"references/training_methods.md",{"path":505,"priority":404},"references/training_patterns.md",{"path":507,"priority":404},"references/troubleshooting.md",{"path":509,"priority":404},"references/unsloth.md",{"path":511,"priority":462},"scripts/convert_to_gguf.py",{"path":513,"priority":462},"scripts/dataset_inspector.py",{"path":515,"priority":462},"scripts/estimate_cost.py",{"path":517,"priority":462},"scripts/hf_benchmarks.py",{"path":519,"priority":462},"scripts/train_dpo_example.py",{"path":521,"priority":462},"scripts/train_grpo_example.py",{"path":523,"priority":462},"scripts/train_sft_example.py",{"path":525,"priority":462},"scripts/unsloth_sft_example.py",{"basePath":291,"description":292,"displayName":293,"installMethods":527,"rationale":528,"selectedPaths":529,"source":278,"sourceLanguage":18,"type":248},{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-local-models/SKILL.md",[530,531,533,535],{"path":288,"priority":272},{"path":532,"priority":404},"references/hardware.md",{"path":534,"priority":404},"references/hub-discovery.md",{"path":536,"priority":404},"references/quantization.md",{"basePath":299,"description":300,"displayName":301,"installMethods":538,"rationale":539,"selectedPaths":540,"source":278,"sourceLanguage":18,"type":248},{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-paper-publisher/SKILL.md",[541,542,544,546,548,550,552,554],{"path":288,"priority":272},{"path":543,"priority":462},"examples/example_usage.md",{"path":545,"priority":404},"references/quick_reference.md",{"path":547,"priority":462},"scripts/paper_manager.py",{"path":549,"priority":462},"templates/arxiv.md",{"path":551,"priority":462},"templates/ml-report.md",{"path":553,"priority":462},"templates/modern.md",{"path":555,"priority":462},"templates/standard.md",{"basePath":307,"description":557,"displayName":309,"installMethods":558,"rationale":559,"selectedPaths":560,"source":278,"sourceLanguage":18,"type":248},"Look up and read Hugging Face paper pages in markdown, and use the papers API for structured metadata such as authors, linked models/datasets/spaces, Github repo and project page. Use when the user shares a Hugging Face paper page URL, an arXiv URL or ID, or asks to summarize, explain, or analyze an AI research paper.",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-papers/SKILL.md",[561],{"path":288,"priority":272},{"basePath":355,"description":563,"displayName":357,"installMethods":564,"rationale":565,"selectedPaths":566,"source":278,"sourceLanguage":18,"type":248},"Use this skill when the user wants to build tool/scripts or achieve a task where using data from the Hugging Face API would help. This is especially useful when chaining or combining API calls or the task will be repeated/automated. This Skill creates a reusable script to fetch, enrich or process data.",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-tool-builder/SKILL.md",[567,568,570,572,574,576,578,580],{"path":288,"priority":272},{"path":569,"priority":404},"references/baseline_hf_api.py",{"path":571,"priority":404},"references/baseline_hf_api.sh",{"path":573,"priority":404},"references/baseline_hf_api.tsx",{"path":575,"priority":404},"references/find_models_by_paper.sh",{"path":577,"priority":404},"references/hf_enrich_models.sh",{"path":579,"priority":404},"references/hf_model_card_frontmatter.sh",{"path":581,"priority":404},"references/hf_model_papers_auth.sh",{"basePath":339,"description":583,"displayName":341,"installMethods":584,"rationale":585,"selectedPaths":586,"source":278,"sourceLanguage":18,"type":248},"Track and visualize ML training experiments with Trackio. Use when logging metrics during training (Python API), firing alerts for training diagnostics, or retrieving/analyzing logged metrics (CLI). Supports real-time dashboard visualization, alerts with webhooks, HF Space syncing, and JSON output for automation.",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-trackio/SKILL.md",[587,588,590,592],{"path":288,"priority":272},{"path":589,"priority":404},"references/alerts.md",{"path":591,"priority":404},"references/logging_metrics.md",{"path":593,"priority":404},"references/retrieving_metrics.md",{"basePath":379,"description":595,"displayName":381,"installMethods":596,"rationale":597,"selectedPaths":598,"source":278,"sourceLanguage":18,"type":248},"Trains and fine-tunes vision models for object detection (D-FINE, RT-DETR v2, DETR, YOLOS), image classification (timm models — MobileNetV3, MobileViT, ResNet, ViT/DINOv3 — plus any Transformers classifier), and SAM/SAM2 segmentation using Hugging Face Transformers on Hugging Face Jobs cloud GPUs. Covers COCO-format dataset preparation, Albumentations augmentation, mAP/mAR evaluation, accuracy metrics, SAM segmentation with bbox/point prompts, DiceCE loss, hardware selection, cost estimation, Trackio monitoring, and Hub persistence. Use when users mention training object detection, image classification, SAM, SAM2, segmentation, image matting, DETR, D-FINE, RT-DETR, ViT, timm, MobileNet, ResNet, bounding box models, or fine-tuning vision models on Hugging Face Jobs.",{"claudeCode":12},"SKILL.md frontmatter at skills/huggingface-vision-trainer/SKILL.md",[599,600,602,603,605,607,608,610,611,612,614,616],{"path":288,"priority":272},{"path":601,"priority":404},"references/finetune_sam2_trainer.md",{"path":495,"priority":404},{"path":604,"priority":404},"references/image_classification_training_notebook.md",{"path":606,"priority":404},"references/object_detection_training_notebook.md",{"path":499,"priority":404},{"path":609,"priority":404},"references/timm_trainer.md",{"path":513,"priority":462},{"path":515,"priority":462},{"path":613,"priority":462},"scripts/image_classification_training.py",{"path":615,"priority":462},"scripts/object_detection_training.py",{"path":617,"priority":462},"scripts/sam_segmentation_training.py",{"basePath":387,"description":619,"displayName":389,"installMethods":620,"rationale":621,"selectedPaths":622,"source":278,"sourceLanguage":18,"type":248},"Train or fine-tune sentence-transformers models across `SentenceTransformer` (bi-encoder; dense or static embedding model; for retrieval, similarity, clustering, classification, paraphrase mining, dedup, multimodal), `CrossEncoder` (reranker; pair scoring for two-stage retrieval / pair classification), and `SparseEncoder` (SPLADE, sparse embedding model; for learned-sparse retrieval). Covers loss selection, hard-negative mining, evaluators, distillation, LoRA, Matryoshka, and Hugging Face Hub publishing. Use for any sentence-transformers training task.",{"claudeCode":12},"SKILL.md frontmatter at skills/train-sentence-transformers/SKILL.md",[623,624,626,628,630,632,634,635,637,639,641,643,645,647,649,650,652,654,656,658,660,662,664,666,668,670,672,674],{"path":288,"priority":272},{"path":625,"priority":404},"references/base_model_selection.md",{"path":627,"priority":404},"references/dataset_formats.md",{"path":629,"priority":404},"references/evaluators_cross_encoder.md",{"path":631,"priority":404},"references/evaluators_sentence_transformer.md",{"path":633,"priority":404},"references/evaluators_sparse_encoder.md",{"path":493,"priority":404},{"path":636,"priority":404},"references/hf_jobs_execution.md",{"path":638,"priority":404},"references/losses_cross_encoder.md",{"path":640,"priority":404},"references/losses_sentence_transformer.md",{"path":642,"priority":404},"references/losses_sparse_encoder.md",{"path":644,"priority":404},"references/model_architectures.md",{"path":646,"priority":404},"references/prompts_and_instructions.md",{"path":648,"priority":404},"references/training_args.md",{"path":507,"priority":404},{"path":651,"priority":462},"scripts/mine_hard_negatives.py",{"path":653,"priority":462},"scripts/train_cross_encoder_distillation_example.py",{"path":655,"priority":462},"scripts/train_cross_encoder_example.py",{"path":657,"priority":462},"scripts/train_cross_encoder_listwise_example.py",{"path":659,"priority":462},"scripts/train_sentence_transformer_distillation_example.py",{"path":661,"priority":462},"scripts/train_sentence_transformer_example.py",{"path":663,"priority":462},"scripts/train_sentence_transformer_make_multilingual_example.py",{"path":665,"priority":462},"scripts/train_sentence_transformer_matryoshka_example.py",{"path":667,"priority":462},"scripts/train_sentence_transformer_multi_dataset_example.py",{"path":669,"priority":462},"scripts/train_sentence_transformer_static_embedding_example.py",{"path":671,"priority":462},"scripts/train_sentence_transformer_with_lora_example.py",{"path":673,"priority":462},"scripts/train_sparse_encoder_distillation_example.py",{"path":675,"priority":462},"scripts/train_sparse_encoder_example.py",{"basePath":371,"description":677,"displayName":373,"installMethods":678,"rationale":679,"selectedPaths":680,"source":278,"sourceLanguage":18,"type":248},"Use Transformers.js to run state-of-the-art machine learning models directly in JavaScript/TypeScript. Supports NLP (text classification, translation, summarization), computer vision (image classification, object detection), audio (speech recognition, audio classification), and multimodal tasks. Works in browsers and server-side runtimes (Node.js, Bun, Deno) with WebGPU/WASM using pre-trained models from Hugging Face Hub.",{"claudeCode":12},"SKILL.md frontmatter at skills/transformers-js/SKILL.md",[681,682,684,686,688,690,692,694],{"path":288,"priority":272},{"path":683,"priority":404},"references/CACHE.md",{"path":685,"priority":404},"references/CONFIGURATION.md",{"path":687,"priority":404},"references/EXAMPLES.md",{"path":689,"priority":404},"references/MODEL_ARCHITECTURES.md",{"path":691,"priority":404},"references/MODEL_REGISTRY.md",{"path":693,"priority":404},"references/PIPELINE_OPTIONS.md",{"path":695,"priority":404},"references/TEXT_GENERATION.md",{"sources":697},[698],"manual",{"closedIssues90d":234,"description":700,"forks":235,"homepage":701,"license":240,"openIssues90d":236,"pushedAt":237,"readmeSize":232,"stars":238,"topics":702},"Give your agents the power of the Hugging Face ecosystem","https://huggingface.co",[],{"classifiedAt":704,"discoverAt":705,"extractAt":706,"githubAt":706,"updatedAt":704},1778690772996,1778689536128,1778690770714,[216,212,217,213,214,215],{"evaluatedAt":244,"extractAt":709,"updatedAt":244},1778690773482,[],[712,742,771,788,818,847],{"_creationTime":713,"_id":714,"community":715,"display":716,"identity":722,"providers":727,"relations":736,"tags":738,"workflow":739},1778691799740.4775,"k17d0yq6vmmtzk249wz61kpa8n86mqrf",{"reviewCount":8},{"description":717,"installMethods":718,"name":720,"sourceUrl":721},"Use when the user is doing AI/ML work in a scientific domain — biology, chemistry, physics, astronomy, climate, genomics, materials science, medicine, ecology, energy, conservation, engineering, mathematics, scientific reasoning, drug discovery, protein design, weather modeling, theorem proving, single-cell, PDE solving, or anything similar. Hugging Science (huggingscience.co) is a curated catalog of scientific datasets, models, blog posts, and interactive Spaces; the `hugging-science` org on Hugging Face hosts community datasets, models, and demo Spaces. This skill helps you discover the right resource AND actually use it — loading datasets via `datasets`, running models via `transformers` or the HF Inference API, calling Spaces like BoltzGen via `gradio_client`, and citing blog posts for methodology. Trigger this skill whenever a user mentions a scientific ML task, asks for \"a dataset/model for X\" where X is a scientific topic, wants to fine-tune on scientific data, asks about protein / molecule / genome / climate / materials / astronomy / pathology / weather ML, or needs AI tools for research — even if they never say \"Hugging Science\" explicitly. The catalog is purpose-built for LLM agents (it ships an `llms-full.txt`); prefer it over generic web search for these tasks.",{"claudeCode":719},"K-Dense-AI/claude-scientific-skills","Hugging Science","https://github.com/K-Dense-AI/claude-scientific-skills",{"basePath":723,"githubOwner":724,"githubRepo":725,"locale":18,"slug":726,"type":248},"scientific-skills/hugging-science","K-Dense-AI","claude-scientific-skills","hugging-science",{"evaluate":728,"extract":734},{"promptVersionExtension":205,"promptVersionScoring":206,"score":209,"tags":729,"targetMarket":218,"tier":219},[212,730,216,215,731,732,733],"science","research","ml","discovery",{"commitSha":254,"license":735},"MIT",{"repoId":737},"kd79rphh5gexy91xmpxc05h5mh86mm9r",[216,733,212,732,215,731,730],{"evaluatedAt":740,"extractAt":741,"updatedAt":740},1778692838166,1778691799740,{"_creationTime":743,"_id":744,"community":745,"display":746,"identity":752,"providers":756,"relations":765,"tags":767,"workflow":768},1778687399826.0234,"k1728fhz3xvt5de6ck3kd5gbsh86ngnm",{"reviewCount":8},{"description":747,"installMethods":748,"name":750,"sourceUrl":751},"Creates, manages, and queries Arize datasets and examples. Covers dataset CRUD, appending examples, exporting data, and file-based dataset creation using the ax CLI. Use when the user needs test data, evaluation examples, or mentions create dataset, list datasets, export dataset, append examples, dataset version, golden dataset, or test set.",{"claudeCode":749},"github/awesome-copilot","arize-dataset","https://github.com/github/awesome-copilot",{"basePath":753,"githubOwner":754,"githubRepo":755,"locale":18,"slug":750,"type":248},"skills/arize-dataset","github","awesome-copilot",{"evaluate":757,"extract":764},{"promptVersionExtension":205,"promptVersionScoring":206,"score":758,"tags":759,"targetMarket":218,"tier":219},100,[760,761,762,214,216,763],"arize","data-management","cli","examples",{"commitSha":254},{"repoId":766},"kd7dsmv976w8rtkqnjjfdtfgks86nnmw",[760,762,761,216,763,214],{"evaluatedAt":769,"extractAt":770,"updatedAt":769},1778689111827,1778687399826,{"_creationTime":772,"_id":773,"community":774,"display":775,"identity":777,"providers":778,"relations":783,"tags":785,"workflow":786},1778690773482.4866,"k17a3mmgvm5hj49twj487hp64186n2qa",{"reviewCount":8},{"description":443,"installMethods":776,"name":333,"sourceUrl":14},{"claudeCode":12},{"basePath":331,"githubOwner":212,"githubRepo":247,"locale":18,"slug":333,"type":248},{"evaluate":779,"extract":782},{"promptVersionExtension":205,"promptVersionScoring":206,"score":758,"tags":780,"targetMarket":218,"tier":219},[762,212,214,761,781],"model-management",{"commitSha":254},{"parentExtensionId":784,"repoId":256},"k175g1spb5757qt4tnj9cktcn986mshy",[762,761,212,214,781],{"evaluatedAt":787,"extractAt":709,"updatedAt":787},1778691223210,{"_creationTime":789,"_id":790,"community":791,"display":792,"identity":798,"providers":803,"relations":812,"tags":814,"workflow":815},1778685991755.708,"k17eak6qjys6kns9c25d40q9kn86n7g2",{"reviewCount":8},{"description":793,"installMethods":794,"name":796,"sourceUrl":797},"Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.",{"claudeCode":795},"davila7/claude-code-templates","huggingface-accelerate","https://github.com/davila7/claude-code-templates",{"basePath":799,"githubOwner":800,"githubRepo":801,"locale":18,"slug":802,"type":248},"cli-tool/components/skills/ai-research/distributed-training-accelerate","davila7","claude-code-templates","distributed-training-accelerate",{"evaluate":804,"extract":811},{"promptVersionExtension":205,"promptVersionScoring":206,"score":805,"tags":806,"targetMarket":218,"tier":219},99,[807,808,809,214,212,810],"distributed-training","pytorch","deep-learning","performance",{"commitSha":254},{"repoId":813},"kd71fzn4s7r0269fkw47wt670n86ndz0",[809,807,212,214,810,808],{"evaluatedAt":816,"extractAt":817,"updatedAt":816},1778686851352,1778685991755,{"_creationTime":819,"_id":820,"community":821,"display":822,"identity":828,"providers":833,"relations":840,"tags":843,"workflow":844},1778696691708.329,"k170yjw596k3cg892bq130jyd186mhyz",{"reviewCount":8},{"description":823,"installMethods":824,"name":826,"sourceUrl":827},"Validate a Claude Code plugin structure, frontmatter, and MCP tool references",{"claudeCode":825},"ruvnet/ruflo","Validate Plugin","https://github.com/ruvnet/ruflo",{"basePath":829,"githubOwner":830,"githubRepo":831,"locale":18,"slug":832,"type":248},"plugins/ruflo-plugin-creator/skills/validate-plugin","ruvnet","ruflo","validate-plugin",{"evaluate":834,"extract":839},{"promptVersionExtension":205,"promptVersionScoring":206,"score":758,"tags":835,"targetMarket":218,"tier":219},[836,289,837,838,213],"validation","developer-tools","claude-code",{"commitSha":254,"license":735},{"parentExtensionId":841,"repoId":842},"k17f4y1y2y777p7zrxxhbnf03n86mr5j","kd7ed28gj8n0y3msk5dzrp05zs86nqtc",[838,837,213,289,836],{"evaluatedAt":845,"extractAt":846,"updatedAt":845},1778701241052,1778696691708,{"_creationTime":848,"_id":849,"community":850,"display":851,"identity":855,"providers":858,"relations":867,"tags":869,"workflow":870},1778696691708.3218,"k17eqpa3h27h177b037g2e8m8n86ndwb",{"reviewCount":8},{"description":852,"installMethods":853,"name":854,"sourceUrl":827},"Track long-horizon objectives across multiple sessions with milestone checkpoints, progress persistence, and drift detection",{"claudeCode":825},"Horizon Track",{"basePath":856,"githubOwner":830,"githubRepo":831,"locale":18,"slug":857,"type":248},"plugins/ruflo-goals/skills/horizon-track","horizon-track",{"evaluate":859,"extract":866},{"promptVersionExtension":205,"promptVersionScoring":206,"score":758,"tags":860,"targetMarket":218,"tier":219},[861,862,863,864,865,213],"objective-tracking","project-management","long-term-goals","persistence","session-management",{"commitSha":254,"license":735},{"parentExtensionId":868,"repoId":842},"k17bh7m6sv83frxqdhrd00vzcx86n03e",[863,213,861,864,862,865],{"evaluatedAt":871,"extractAt":846,"updatedAt":871},1778700614118]